EPTMy ZvP has gone to crap in the past couple weeks as Zerg players watched Stephano stomp all over Protosses, and it seems that NA Zergs have finally figured out the proper timings on a 3 hatch speed roach build.
Zergs used to do silly things like 4 hatch before gas and drone straight to 80 with a 9 minute roach warren. At the time, they were thinking that all they had to worry about a 10 minute 7 gate, and we punished them with cute +1 zealot timings at 8 minutes.
Then they got safer with a 7 minute roach warren, but they'd delay gas and lair, so we'd take a third behind some light pressure--perhaps with a void ray to stop slow roaches from countering--and we'd set up a defensive position at our third in the timing window before they'd get a proper attack force with roach speed.
Then Stephano showed them that you can get 2 gas at 6 minutes, a roach warren and evo chamber at 7 minutes with a lair quickly afterwards, and you'll be totally safe against the early pressure timings and still hit 60 drones 8 minutes. Then simply pump roaches and you'll have speed and a 160 supply army by 11 minutes, and you'll be maxed on +1 speed roaches by 12 minutes.
Then you box your roaches, hit A and click on the Protoss's base on the minimap while you continue injecting and hitting SR. And you win every time cuz you're up 80 supply.
It looks a little like this:
http://drop.sc/134488
Mind you, I didn't play well because I was tilting after losing like 5 in a row to the same build, but just watch the Zerg side and tell me what P is supposed to do. With great forcefields, I could have survived the first 200/200, but I still would have had no shot to win the game. Just imagine trying to take a third against that.
Or you can check out a game of Stephano doing it. This is a good one:
So I really have no clue how to beat it. I've been trying everything. In desperation, I've been trying FFE -> 1 gate robo -> 7 minute third and trying to hold the fort with +1 zealots and immortals. The idea is that roach speed doesn't come until 11 minutes, so if I take the third fast enough, maybe I can contest Z's production in the 10-15 minute period. +1 zealots with immortals can handle lings + slow roaches pretty efficiently. Here's a rep of me trying it and winning. I assure you there have been plenty of losses as well.
http://drop.sc/134487
What are your ideas? 200/200 speed roaches at 12 minutes with no vulnerable window to exploit is.....rough.
Note: this isn't a [H] thread--I can see the stuff I did wrong in my replays. I want to talk general strategy, not that I missed a hatch block, accidentally cut some probes, didn't forcefield well, and probably could have fared a little better with a warp prism instead of an observer. I think it's pretty clear that there needs to be some major strategic shifts rather than incremental execution improvements.
______________________________________________________________________
The best proposed solution I've seen in the thread so far is FFE to 4 gate robo to third base before 8 minutes with plenty of sentries. The idea is that by skipping the opportunity to pressure off 2 bases and instead expand quickly to your third, you gain enough production to survive the 12 minute roach timing.
The following replay was submitted showing the build:
http://drop.sc/134585
Ranged, a high level Protoss player, commented that he has some experience with the build:
On March 16 2012 19:01 RaNgeD wrote:
I've been using this build for a while, and instead of getting an obs I get hallucination. That way it's possible to even have an extra immortal out if the zerg tries to hit a timing and bust the 3rd. You're getting a cannon at the 3rd, so unless you are really worried about burrow, I don't see a problem with cutting this corner. Also with hallucination you can be scouting for a muta switch, and you can place a Halluc pheonix to follow his roach army. More vision = better forcefields.
I've been using this build for a while, and instead of getting an obs I get hallucination. That way it's possible to even have an extra immortal out if the zerg tries to hit a timing and bust the 3rd. You're getting a cannon at the 3rd, so unless you are really worried about burrow, I don't see a problem with cutting this corner. Also with hallucination you can be scouting for a muta switch, and you can place a Halluc pheonix to follow his roach army. More vision = better forcefields.
_______________________________________________________________________
Here's some more good content that might have been missed by some since it's buried in a rather long thread:
On March 20 2012 18:05 RaNgeD wrote:
Thanks for sharing. This may sound obvious to some, but i'm of the opinion that to play against zerg you have two choices:
1) Aggressively expand in order to keep up with their economy; or 2) Do timings or harass to kill them or at the very least, slow them down before hive tech. If you choose to play passively against zerg I think you need to be playing aggressive economically.
That being said, I really like what SaSe was doing in Game 1 against ret. I don't know how viable it is, and I don't know how zergs will react to a fast 3rd like that once they've played against it a few times and have had the chance to sit down and study it out a bit. But the general idea is amazing to me.
In my last 100-200 pvz games I've done the fast 3rd off of 4 gates/robo like mentioned earlier in this thread, and from my experience I can say this:
The times when zerg decide to max on roaches at 12 minutes and kill me, 50% of the time I die. And the games that I do win, it was usually that the zerg made some critical mistake, like letting me cut off 1/2 of 1/4 of his roaches and kill them for free. I've made a lot of adjustments to the build to try and survive that 12 minute wave of 200/200 roaches, like getting hallucination instead of an obs, that way i'm able to get an extra immortal out, and just use cannons and detection. Keep in mind that these are ladder games; if I could find some customs against good zergs it would be nice, but for the most part, these players aren't even near the calibre of stephano. I can't say for sure, but i'm coming to the conclusion that I think the 3rd base with this current build is too slow. The 3rd base doesn't kick in enough before you are over-run with roaches.
Another thing I wanted to say here. I don't find non-FFE builds very good at all. I used them for the first year and eventually came to the conclusion that FFE is much much better. Maybe that comes down to my style or something, but I don't really see how doing a less economic build is going to solve this issue. Even if it did, with gate expands you run into a completely different set of problems.
What SaSe is doing is really good, even if it doesn't end up working in the long run. We can sit here and do the same builds over and over again, expecting different results; OR, we can make new builds, study and refine them, and hopefully stumble upon a better way to play. I like the latter.
Thanks for sharing. This may sound obvious to some, but i'm of the opinion that to play against zerg you have two choices:
1) Aggressively expand in order to keep up with their economy; or 2) Do timings or harass to kill them or at the very least, slow them down before hive tech. If you choose to play passively against zerg I think you need to be playing aggressive economically.
That being said, I really like what SaSe was doing in Game 1 against ret. I don't know how viable it is, and I don't know how zergs will react to a fast 3rd like that once they've played against it a few times and have had the chance to sit down and study it out a bit. But the general idea is amazing to me.
In my last 100-200 pvz games I've done the fast 3rd off of 4 gates/robo like mentioned earlier in this thread, and from my experience I can say this:
The times when zerg decide to max on roaches at 12 minutes and kill me, 50% of the time I die. And the games that I do win, it was usually that the zerg made some critical mistake, like letting me cut off 1/2 of 1/4 of his roaches and kill them for free. I've made a lot of adjustments to the build to try and survive that 12 minute wave of 200/200 roaches, like getting hallucination instead of an obs, that way i'm able to get an extra immortal out, and just use cannons and detection. Keep in mind that these are ladder games; if I could find some customs against good zergs it would be nice, but for the most part, these players aren't even near the calibre of stephano. I can't say for sure, but i'm coming to the conclusion that I think the 3rd base with this current build is too slow. The 3rd base doesn't kick in enough before you are over-run with roaches.
Another thing I wanted to say here. I don't find non-FFE builds very good at all. I used them for the first year and eventually came to the conclusion that FFE is much much better. Maybe that comes down to my style or something, but I don't really see how doing a less economic build is going to solve this issue. Even if it did, with gate expands you run into a completely different set of problems.
What SaSe is doing is really good, even if it doesn't end up working in the long run. We can sit here and do the same builds over and over again, expecting different results; OR, we can make new builds, study and refine them, and hopefully stumble upon a better way to play. I like the latter.
Discussing progressively earlier thirds inspired me to do a little math. Summing up the costs associated with the attack and the defense might give some guidelines for figuring out what sorts of builds might work. The numbers aren't supposed to incorporate everything--they're just a starting point for focusing ideas.
On March 21 2012 01:52 kcdc wrote:
I think that Sase's opening is really interesting as well, but I think even that fast of a third will wind up well behind in army size. First, a stalker roughly equals a roach in combat strength, but costs twice as much. Second, Z has to spend substantially less on infrastructure for the 3-base roach max compared to a 3-base Protoss defense.
SuperRoach (counting drone in building cost):
pool, warren: 450 minerals
3 hatches (2 expansion, 1 macro): 1050 minerals
3 queens: 450 minerals
evolution chamber: 125 minerals
lair: 150/100
ling speed, roach speed and +1 missile: 300/300
4 gas: 300 minerals
Total: 2825/400
SuperRoach Defense:
2 nexii: 800 minerals
forge, core: 300 minerals
4 cannons minimum: 600 minerals
WG and +1 weapons: 150/150
8 gates (minimum, you'll want more on 3-base econ): 1200
robo: 200/100
5 gas: 375 minerals
twilight (necessary before 12 minutes): 150/100
observer (or hallucination, but we'll count the cheaper one): 25/75
Total: 3800/425
So ~1000 minerals is the minimum infrastructure difference as I see it. If you want some extra cannons or blink or a robo support bay or some extra gates, it's going to add up pretty fast.
Factor in that you'll never fully catch Z's economy even with a third nexus before core and that roaches cost about half as much per supply as your composition does, and it becomes pretty clear that if P wants to take a fast third, Z can have a 70 supply lead at 12 minutes if they want it.
So I think that while it's tempting to ask, "How do I match Z's economy and army size?" the more important question may be, "How do I defend 2+ attack paths when I'm down 70 supply?" I think the answer to the latter question is going to rest heavily on forcefields, so I wonder if Sase's delayed core build is such a good choice.
I think that Sase's opening is really interesting as well, but I think even that fast of a third will wind up well behind in army size. First, a stalker roughly equals a roach in combat strength, but costs twice as much. Second, Z has to spend substantially less on infrastructure for the 3-base roach max compared to a 3-base Protoss defense.
SuperRoach (counting drone in building cost):
pool, warren: 450 minerals
3 hatches (2 expansion, 1 macro): 1050 minerals
3 queens: 450 minerals
evolution chamber: 125 minerals
lair: 150/100
ling speed, roach speed and +1 missile: 300/300
4 gas: 300 minerals
Total: 2825/400
SuperRoach Defense:
2 nexii: 800 minerals
forge, core: 300 minerals
4 cannons minimum: 600 minerals
WG and +1 weapons: 150/150
8 gates (minimum, you'll want more on 3-base econ): 1200
robo: 200/100
5 gas: 375 minerals
twilight (necessary before 12 minutes): 150/100
observer (or hallucination, but we'll count the cheaper one): 25/75
Total: 3800/425
So ~1000 minerals is the minimum infrastructure difference as I see it. If you want some extra cannons or blink or a robo support bay or some extra gates, it's going to add up pretty fast.
Factor in that you'll never fully catch Z's economy even with a third nexus before core and that roaches cost about half as much per supply as your composition does, and it becomes pretty clear that if P wants to take a fast third, Z can have a 70 supply lead at 12 minutes if they want it.
So I think that while it's tempting to ask, "How do I match Z's economy and army size?" the more important question may be, "How do I defend 2+ attack paths when I'm down 70 supply?" I think the answer to the latter question is going to rest heavily on forcefields, so I wonder if Sase's delayed core build is such a good choice.
I'm rethinking that conclusion that Z can be ahead by 70 supply no matter what you do. Ranged and I, among others, have been experimenting with taking a third between 6:30 and 8:00, and it seems possible to hit 140+ supply by 12 minutes which is when Z usually queues up the production round that pushes his supply to 200/200. Of course, that production round typically doesn't hit P's base for another minute, so the force that's at P's door at 12:00 is really a 170 supply army with 30 supply en route. With WG, P doesn't lose suffer the production time+travel distance penalty, so the 140 supply is a hard number. So with a very optimized build and near-perfect play, P might be able to find themselves down only ~30 or so supply at the point of attack.
